> As another sysadmin I'd agree with this. Kernel crashes nearly always > boil down to bad hardware, or bugs in the kernel and/or it's drivers. > > For the OP: > > Without knowing what your hardware, OS or location is...
supermicro-somthing, with two dual xeon, 3gb of RAM, gentoo linux, germany ;-) > > Thoeretically a decent modern kernel "shouldn't" be able to be crashed > by userland software no matter how buggy - although they can still > cause resource starvation etc. But Python code is another layer > removed from the kernel anyway - eg theoretically buggy Python code > "shouldn't" be able to even crash Python itself (but hey they could > happen too). > that´s what i´m thinking, too... i couldn´t imagine that this TG-app made the server go away. especially as the server stopped most of the time in the middle of the night, when there are no calls to the app. > I would be inclined to look at the hardware myself - it ends up being > the problem more often than the kernel or drivers. Apparently you've > tested the RAM (I assume it isn't ECC RAM?), but the only way to know > for sure is to swap out part/all of it. Dodgy power supplies can also > cause all sorts of random issues. Can you reproduce the problem on > other hardware? > yes, we´ve tested the ram but without any result - the test (the vendor recommended) ran through without error. i´m not quite sure if it´s ecc ram or now - can i find it out somehow when running? or can only a look at the hardware tell me? i reanimated an old server machine today and installed exactly the same application there as well. i will now make it run some days - for sure without crashing... > RAM, power supplies and cooling fans are the top culprits for weird > random failures. > > Also apparently the crashes happen overnight at low loads. Maybe it is > environmental? Is your server just sitting in an office or cupboard > somewhere? Does the AC shut off at night? The server could be > overheating. > the server is in a separate room, which does not have AC, he´s in a rack with some other servers, with fans starting when the in-rack air- temperature reaches 35° C (95° F), but most of the time they are not running. i think if they would overheat they would rather do so during the day when there´s load on them. i test now everything with the old hardware, and if it´s not crashing, the cendor will have to check the server´s hardware... :-(((( --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/turbogears?hl=en -~----------~----~----~----~------~----~------~--~---
